The ethanol withdrawal syndrome, which is partially characterized by an increased activity of the noradrenergic system, is at present most commonly treated with diazepam or chlordiazepoxide, both conventional benzodiazepines. Alprazolam, a new benzodiazepine, has been demonstrated to be successful in the pharmacotherapy of depression and anxiety disorders, in contrast to the conventional benzodiazepines. Alprazolam may have a particularly potent inhibitory action on the noradrenergic system. It can, therefore, be postulated that alprazolam may be an effective and specific treatment for the ethanol withdrawal syndrome. Clonidine, a conventional antihypertensive, has been used to successfully treat withdrawal from the opiates, and most recently, nicotine and alcohol. This study will compare the effects of alprazolam, clonidine, diazepam, and placebo on: 1) the signs and symptoms of the ethanol withdrawal syndrome, and 2) the noradrenergic overactivity of the ethanol withdrawal syndrome. Noradrenergic activity will be evaluated by determinations of cerebrospinal fluid (CSF), and plasma, catecholamines and their metabolites, plasma norepinephrine laying and standing, and lymphocyte B-adrenergic receptor sites. Also, changes in CSF pH have been reported during ethanol withdrawal. We will, therefore, evaluate CSF pH both during acute withdrawal and following three weeks of abstinence.